Cardiac Troponin I and Cardiovascular Risk in Patients With Chronic Obstructive Pulmonary Disease

Background Patients with chronic obstructive pulmonary disease (COPD) have increased risk of cardiovascular events. Objectives This study evaluated the association between high-sensitivity cardiac troponin I concentration and cardiovascular events in patients with COPD and heightened cardiovascular risk. Methods In a double-blind randomized controlled trial, 16,485 patients with COPD and cardiovascular disease or risk factors were randomized to once daily inhaled placebo, fluticasone furoate (100 μg), vilanterol (25 μg), or their combination. Plasma high-sensitivity cardiac troponin I concentrations were measured in a subgroup of 1,599 patients. Outcomes were on-treatment cardiovascular events and COPD exacerbations over a median of 18 months, and cardiovascular death over a median of 27 months. Results Baseline plasma cardiac troponin I concentrations were above the limit of detection (1.2 ng/l) in 1,542 (96%) patients. Concentrations were unaffected by inhaled therapies at 3 months (p > 0.05). Compared with the lowest quintile (cardiac troponin <2.3 ng/l), patients in the highest quintile (≥7.7 ng/l) were at greater risk of cardiovascular events (hazard ratio [HR] 3.7; 95% confidence interval [CI]: 1.3 to 10.1; p = 0.012) and cardiovascular death (HR: 20.1; 95% CI: 2.4 to 165.2; p = 0.005) after adjustment for risk factors. By contrast, there were no differences in exacerbations between quintiles (HR: 1.1; 95% CI: 0.8 to 1.5; p = 0.548). Conclusions In patients with COPD and heightened cardiovascular risk, plasma cardiac troponin I concentrations are a specific and major indicator of future cardiovascular events and cardiovascular death. Inhaled therapies did not affect cardiac troponin I concentrations consistent with their neutral effect on mortality and cardiovascular outcomes. (Study to Evaluate the Effect of Fluticasone Furoate/Vilanterol on Survival in Subjects With Chronic Obstructive Pulmonary Disease [SUMMIT]; NCT01313676)

C ardiovascular (CV) disease, including ischemic heart disease and stroke, accounts for 1 in 4 deaths globally and is increasing in prevalence (1). Despite recent advances in understanding risk factors and therapeutic interventions, atherosclerotic events remain unacceptably common.
Residual risk is particularly high among patients with proinflammatory comorbidities, such as chronic obstructive pulmonary disease (COPD) (2). In some cases, it remains unclear whether it is the disease process itself, or the off-target effects of the pharmacological treatments, that contribute to this elevated risk (3).
Notwithstanding prior major advances, future clinical trial conduct is hampered by several important and increasing challenges. It is well recognized that clinical trial participants represent a relatively low-risk subset of the real-world patient population.
Consequently, modest event rates necessitate large and costly trials in order to demonstrate treatment efficacy. Conversely, this low event rate creates the potential for researchers to fail to recognize CV harms  (5)(6)(7)(8)(9)(10)(11). Furthermore, plasma cardiac troponin I concentrations measured by a high-sensitivity assay have recently been shown to be modifiable, with statin-induced reductions in cardiac troponin I proving a more powerful indicator of treatment efficacy than changes in serum cholesterol (12). The role of serial testing with high-sensitivity cardiac troponin I to predict the effect of other therapies on CV outcomes has to date been unexplored in patients with more diverse multimorbid conditions. SUMMIT (Study to Understand Mortality and MorbidITy) (13,14) assessed the efficacy and safety of inhaled corticosteroids and longacting beta-agonists (LABAs) in 16,485 patients with COPD and heightened CV risk. This was a multimorbid population with interventions that could have both benefit (15) and harm (16). The present study reports post hoc analyses aiming to determine whether plasma high-sensitivity cardiac troponin I concentrations can stratify CV risk, be modified by inhaled corticosteroids (ICS) and bronchodilators, and predict outcomes within the context of SUMMIT.  (13,14).

METHODS
In brief, eligible participants included current or former smokers ($10 pack-years) between the ages of 40 and 80 years, with a history of COPD and a post-bronchodilator forced expiratory volume in 1 s (FEV 1 ) $50% and #70% of the predicted value, a ratio of post-bronchodilator FEV 1  ENDPOINTS. In addition to the primary endpoint of all-cause mortality by intention-to-treat analysis, the secondary CV endpoint was time to first-ontreatment CV event comprising CV death, myocardial infarction, stroke, unstable angina, and transient ischemic attack (17). Categorization of the cause of HIGH-SENSITIVITY CARDIAC TROPONIN I. Venous blood samples were obtained before randomization and at 3 months. Blood was processed and plasma stored at À80 C until analyzed. As previously described, before analysis, samples were thawed and underwent centrifugation twice (3,000 relative centrifugal force for 10 min) according to the manufacturer's instructions to ensure homogeneity (12). Plasma highsensitivity cardiac troponin I concentrations were measured at a single site using the ARCHITECT STAT high-sensitive cardiac troponin I assay (Abbott Laboratories, Abbott Park, Illinois), which has a limit of detection of 1.2 ng/l, coefficient of variation <10% at 4.7 ng/l, and sex-specific 99th percentile upper reference limits of 16 and 34 ng/l in women and men, respectively (18,19). where the best fitting model was selected from a variety of polynomial or logarithmic models using 2-term fractional polynomials (20).
Previous reports have identified that adverse CV outcomes are associated with plasma troponin I concentrations $5 ng/l (12,21). To explore this association further, and investigate whether the predictive value of this threshold could be applied to the SUMMIT population, patients were grouped into those who had plasma concentrations <5 ng/l at both baseline and 3 months, and those with a concentration $5 ng/l at either baseline or 3 months.
Scientific oversight of the trial was provided by a steering committee composed of academic experts and employees from GlaxoSmithKline, who were collectively responsible for the study design and analysis, and for the review and interpretation of the data. This study is registered with ClinicalTrials.gov (NCT01313676).

RESULTS
The study population and principal findings of the SUMMIT study have previously been described (14). The patient characteristics in the biomarker substudy population were broadly similar to the overall SUMMIT population, except that those in the biomarker population were more likely to be female, have a higher body mass index, have fewer previous COPD exacerbations, and have differences in CV history and CV therapy ( Table 1). Participants were stratified into quintiles by plasma cardiac troponin I concentration from samples obtained before randomization. Compared with the lowest Compared with the lowest quintile, participants in the highest quintile were at greater risk of experiencing a CV composite event ( Figure 1A). This difference persisted after adjustment for confounding variables including CV risk factors (    Abbreviations as in Table 1. Adamson et al.     Adamson et al. The association between baseline high-sensitivity cardiac troponin I and CV death (orange) and COPD exacerbations (blue) was examined using cardiac troponin as a continuous variable. Hazard ratios are compared with the median troponin concentration in the first quintile (1.7 ng/l) and are adjusted for age, sex, previous myocardial infarction, hypertension, and exacerbation history. Shaded areas represent 95% confidence intervals. COPD ¼ chronic obstructive pulmonary disease; CV ¼ cardiovascular.
JACC VOL. 72, NO. 10,2018 Adamson et al. low-risk groups. This cutpoint of risk is remarkably consistent with previous descriptions (11,12,21,(39)(40)(41), and supports the concept of a threshold value above which event rates rise substantially. This threshold is well below the 99th percentile upper reference limit for this assay used for the diagnosis of myocardial infarction, and adoption of troponin testing for CV risk stratification will require additional guidance for clinicians. It is important to note that the distribution of cardiac troponin is highly skewed in reference populations (19), with the upper reference limit increased markedly by a small number of outliers who are likely to have subclinical disease. As such, the cumulative evidence from studies evaluating prognosis rather than diagnosis suggest that the threshold to define low risk (the true normal, perhaps) is at the much lower concentration of 5 ng/l (4). The magnitude of increased risk we identified across troponin quintiles is similar to that seen in both primary and secondary CV disease prevention populations, reinforcing its broad applicability for prognostic stratification (7,11,12). Importantly, this threshold is above the 10% coefficient of variation for this assay, and could be used to guide treatment   troponin testing using this assay (11,12,21,39,40).
Although future research studies are required to quantify the effects of analytical and biological variability through repeated sampling in this patient population, misclassification could be addressed in clinical practice through the use of serial troponin measurements over consecutive clinic visits. This would be analogous to the recommendation in clinical guidelines to undertake repeated blood pressure measurements on separate occasions before conferring a diagnosis of hypertension.

CONCLUSIONS
In patients with combined respiratory and CV diseases, high-sensitivity plasma cardiac troponin I concentration is a prognostic marker that is specific to